1. Field of the Invention
The present invention relates to a powerline communication system, and more particularly to a packet communication method of a powerline communication system.
2. Brief Description of the Prior Art
A description will be made on a packet communication method of a powerline communication system in accordance with the prior art below.
A powerline communication is a communication method in which high frequency communication signals of hundereds of and tens of MHz are sent along to a powerline, through which an alternating current having a frequency of 50˜60 Hz is supplied into private houses, but only the high frequency communication signals are received by exclusively used connection equipment. The method as such is a convenient, economical communication network by getting connected to an electric outlet to be used as an external network for an internet connection or local area communication network for an homeLAN without installation of an expensive exclusive powerline or basic network.
In such a powerline, there are many loads (hereinafter referred to as nodes) of home appliances such as TV, VCR, Computer and the like. The powerline has a bus type of a network structure where a plurality of nodes may be in connection with a single medium and it takes a period of time for packet transmission. At this time, there may be a collision phenomenon resulting from the simultaneous transmission of two nodes when the two nodes do not know respective transmission states. In order to solve the aforementioned problem of a multiple connection, there has been disclosed a method named a carrier sense multiple access (CSMA), in which a signal transmission is begun after it is confirmed before the transmission that no channel is in use and no signal is present in the channels. However, this method is not helpful to avoid the collision if two nodes make transmissions at an exactly identical time. The aforementioned phenomenon happens when more than two nodes wait for the time when channels, most of which are occupied by particular nodes, become available for use after completion of communication and all the nodes start transmission at the same time.
In order to prevent a collision, it is necessary to prevent two nodes from waiting together and making a simultaneous transmission. As a method to solve the aforementioned problem, there has been a method in which two nodes waiting for the time to use channels discriminate whether the nodes become available for use and then start a transmission after respective random time delays. At this time, if there is a difference in the time delays, one node firstly uses a channel and the other node starts communication after waiting for the channel to be available again.
However, there have been problems in the aforementioned method of the prior art. Firstly, there may be a collision if the random time delays happen to have an identical value. Secondly, one node is continuously pushed by the other node to start a communication much later than a normal time, so that in most cases communicatin time may exceed its limit to cause disuse of a packet.
To solve the first problem, the number of collisions can be cut down not to make any influence on communication by increasing the range of random time delay values properly to the number of nodes. On the other hand, there has been no particular solution to solve the second problem. What's worse, if three nodes are involved in communication and two nodes are in communication holding most of time slot like a download using a file transfer protocol therebetween, a third node can not get into communication with readiness. The phenomenon described above may be solved when a priority is put to specific nodes. On the contrary, in the worst case that a node having a lower priority is at a disadvantage, it can not be a solution because there may be no communication at all.